Goal
Keep the knowledge graph healthy: reduce orphans, increase meaningful connectivity, and create bridges between clusters.
Quick start
- Define scope: directories + time window (e.g. last 30/90 days).
- Diagnose first, prescribe second: deliver only the top 3–5 structural issues and a task list.
Workflow
Step 1: Scan
- Count notes / links
- Identify hubs and clusters
- Identify outliers (orphans, weak links)
Step 2: Diagnose
Example thresholds:
- Orphan rate: <5% excellent / 5–10% good / 10–20% ok / >20% bad
- Avg degree: >3 excellent / 2–3 good / 1–2 ok / <1 bad
Step 3: Identify issues
- Orphans (0 links)
- Weak links (<2 links)
- One-way links
- Cluster isolation (no bridges)
- Tag/label inconsistency (if you use tags)
Step 4: Prescribe
Prioritize:
- Bridge work (cluster-to-cluster)
- Orphan rescue (high value notes first)
- Index creation / consolidation
Output template
## 🔧 Network Maintenance Report
### Snapshot
- **Total notes**: [N]
- **Total links**: [N]
- **Orphans**: [N] ([%])
### Health score: [0-100]
### 🚨 Top issues (prioritized)
1. **[Issue type]**: [what] → [why it matters] → [fix]
### 📋 Task list
- [ ] [Task 1]
- [ ] [Task 2]
Quality checks
- Metrics → diagnosis → tasks align 1:1
- Tasks are concrete (which notes, what links)
- Priorities are explicit